You are on page 1of 15

Hydrobiologia (2012) 684:161175 DOI 10.



Thermal impact of a nuclear power plant in a coastal area in Southeastern Brazil: effects of heating and physical structure on benthic cover and sh communities
Tatiana Pires Teixeira Leonardo Mitrano Neves jo Francisco Gerson Arau

Received: 22 September 2010 / Revised: 27 August 2011 / Accepted: 23 December 2011 / Published online: 19 January 2012 Springer Science+Business Media B.V. 2012

Abstract The inuence of a nuclear power plants cooling water and physical structure on benthic cover and sh communities were assessed in a coastal area in Southeastern Brazil. We hypothesised that thermal discharges decrease benthic cover and consequently, change the associated rocky reef sh assemblage structure and that physical structure is directly associated with sh richness and diversity. Twelve sites at different distances (close, near and far) from thermal discharge and types of physical structure (low and high) were sampled by visual census. The average surface temperature at the most impacted sites (close) ranged from 30.5 to 31C, while at far sites it ranged from 25.5 to 28.5C. Although thermal inuences have decreased benthic cover, and consequently, decreased sh richness and diversity, we found that in near and far sites that had complex habitat structures (physical and benthic cover) sh communities were unaffected. The greatest abundances of Eucinostomus argenteus, Mugil curema and Sphoeroides greeleyi were associated with the highest temperatures at the most impacted sites. In contrast, Abudefduf saxatilis, Chaetodon striatus, Stegastes fuscus, Diplodus

argenteus and Malacoctenus delalandii were more abundant at high structured sites far from thermal discharges. Our data support the hypothesis that thermal discharge decreases benthic cover, sh richness and diversity but physical structure, when coupled with high diversity and abundant benthic cover, minimised thermal effects on sh communities. Keywords Thermal pollution Structural complexity Habitat Rocky shore shes

Introduction Temperature is a very important ecological parameter that affects almost every aspect of aquatic life. Heat from the cooling water of nuclear power plants changes the biological and ecological components of coastal area systems in an unpredictable manner. These effects depend on the quantity of heated discharge, climate and the biological features of the environment (Schubel et al., 1978; Lardicci et al., 1999; Chou et al., 2004). Depending on the design and the operating units of the power plants, water temperature in efuent sites can increase by as much as 8C (Laws, 1993). Consequently in tropical oceans, seawater temperatures can rise to 30C or higher during the summer. Such high temperatures may approximate or even exceed what resident organisms can tolerate (Jokiel & Coles, 1974; Suresh et al., 1993; Wright et al., 2000). Thus, condenser efuents have the

Handling editor: I. A. Nagelkerken jo (&) T. P. Teixeira L. M. Neves F. G. Arau rio de Ecologia de Peixes, Universidade Federal Laborato dica, Rural do Rio de Janeiro, BR 465, Km 7, Serope RJ 23.890-000, Brazil e-mail:



Hydrobiologia (2012) 684:161175

potential to trigger thermal and chemical stress (e.g. dissolved oxygen depletion, water chlorination) and therefore, may pose environmental problems to the receiving water body (Krishnakumar et al., 1991; Chou et al., 2004). Tropical marine organisms are particularly susceptible to thermal effects because the water temperature in which they live in is generally close to their vital limits (Langford, 1990). Several studies have assessed thermal inuence on sh, mainly in temperate areas. Some of these studies refer to the biological features of single species that have been measured or observed in the eld, such as early development and distribution (Shuter et al., 1985; Madenjian et al., 1986), reproduction (Luksiene m, 1994), and life history (Dembski et al., & Sandstro 2006). Others are restricted to laboratory experiments on thermal tolerance (Bennett & Judd, 1992; Mora & na, 2001; Ospina & Mora, 2004). The effects of Osp warmer temperatures on freshwater species can be both direct and indirect, and furthermore, the magnitude of these effects can range from minor to fully lethal (Verones et al., 2010). Direct effects include increased activity with faster digestion, which leads to greater food demand and disturbed reproduction m, (Sandstrom et al., 1997; Luksiene & Sandstro 1994). Indirect effects are related to changes in food availability, community structure, pathogens, chemical processes (modied oxygen content, increased effects of some pollutants) and competition with other, better adapted species (Mariazzi et al., 1992; Penaz et al. 1999; Beitinger et al., 2000; Contador, 2005; Dembski et al., 2006; Encina et al., 2008; De Vries et al., 2008). However, in tropical coastal systems, the information available on how thermal inuence impacts sh communities is limited (Teixeira et al., 2009). Many sh food sources and shelters (e.g. corals, sponges, macroalgae) are sessile, and will be adversely affected because benthic species are particularly susceptible to thermal discharge effects (Bamber & Spencer, 1984; Bruno et al., 2007). Accordingly, sh may not be able to avoid the indirect effects of thermal pollution, such as decreased benthic cover or even the total lack of benthic cover, because it is used as a feeding resource or for protection. Fish are motile, and most can emigrate to safe areas if temperatures rise above tolerable levels, which would ultimately change their community structure. Moreover, apart from simply leaving an area, rock-dwelling sh may respond

more subtly by changing species distribution (RongQuen et al., 2001). Furthermore, a decrease in habitat complexity due to thermal pollution is also expected to decrease sh richness and benthic cover and increase opportunistic and ephemeral species, thereby changing the population dynamics (Devinny, 1980; Mahadevan, 1980; Verlaque et al., 1981; Bamber & Spencer, 1984; Suresh et al., 1993; Qian et al., 1993; Chou et al., 2004). Besides benthic cover, physical structure is known to inuence sh richness because it forms a complex framework that supports a variety of microhabitats, thus increasing richness when increasing complexity (Roberts & Ormond, 1987; Chabanet et al., 1997; hman & Rajasuriya, 1998). Attributes such as number O of holes and crevices, and rock size have been used as physical descriptors (McCormick, 1994). Gladfelter & Gladfelter (1978), working with coral reef communities in the western Atlantic, suggested that sh abundance increases with physical structure. The spatial distribution of topographical characteristics like rock size, holes and bottom types provide organisms with food, shelter and reproductive grounds (AburtoOpereza & Balart, 2001). Therefore, it is widely accepted that sites with high physical structure and diverse and abundant benthic cover have increased sh richness and diversity. In the tropics, there are few nuclear power stations, scattered throughout Mexico, India, South Africa, Argentina and Brazil, and studies on thermal inuence on sh communities are lacking. In Brazil, there are two nuclear power stations that pump approximately 120 m3 s-1 of cooling water to generate 1,900 MW of power. These power plants use seawater from Ilha Grande Bay, an embayment area in Southeastern Brazil, for cooling. Given that its operation is likely to modify some physical and chemical characteristics of the local rocky shore environment, and consequently its biological features, it provides an interesting opportunity to evaluate thermal effects on sh communities. We tested the hypothesis that species richness, abundance and diversity in sh communities are inuenced by thermal discharges, and that the physical structure of the habitat can minimise such effects because it increases diversity and richness. Our aims were to interpret the overall variation in the structure of a rocky reef sh assemblage across a thermal gradient, and to study the relationship between physical structure and sh richness and abundance. We also investigated the indirect effects of water


Hydrobiologia (2012) 684:161175


warming on sh as a result of the benthic cover modications.

Materials and methods Study area

because they were located in a small (2 km2) embayment, and the far sites were located outside of the inuenced area, *3.8 km and 9.0 km from the discharge outfall. We chose rocky shore sites in calm waters that were not near sandy beaches or estuarine areas. Physical structure

This study was conducted along a rocky shore in Ilha Grande Bay, close to the water discharge of the Brazilian Nuclear Power Plant (BNPP), which is comprised of two power plant units. Thermal discharge in this area is a local anomaly that can reach some square kilometres from the outfall (Lucca et al., 2005). This thermal efuent is known to produce an increase of up to 8C in the area of discharge compared to the adjacent area all year round (Bandeira et al., 2003). Twelve sites were selected and classied according to distance from the thermal discharge: four close (\200 m), four near (2001,500 m) and four far ([1,500 m) from the impact source (Fig. 1). The close sites were directly affected by the thermal outfall, the near sites were still affected by outfall
Fig. 1 Map of the study area with indication of the sampling sites coded by distance from the outfall (close, near, far) and physical structure (low, high): CL close low, CH close high, NL near low, NH near high, FL far low, FH far high

Sites also varied in physical structure. Each distance level had two low structured and two high structured sites. The physical structure of sites was assigned a priori as either high or low based on previous observations of topographical complexity and substratum diversity (Teixeira personal observation). This qualitative assignment was subsequently conrmed with a more detailed quantitative analysis based on a physical structure index. A housed digital camera was mounted onto a 1-m2 polyvinyl chloride (PVC) photo quadrat framer. Thirty photographs taken in each sampling plot of 90 m2 were used to quantify the variables describing physical structure following Chapman et al. (1995). Photographs were downloaded to a personal computer, renamed with a unique site



Hydrobiologia (2012) 684:161175

code, and then cropped and colour autocorrected with image-editing software. The number of each descriptor was counted in each photograph (number per square metre). Each site was analysed based on photographs to quantify the following physical descriptors: rugosity, refuge size categories (holes and crevices), substratum, height and hard substratum. The substratum prole across the 90 m2 transect was examined to assign a rugosity score (in this study substratum refers to sand, boulders, bare rocks and any articial structures). Flat sandy areas were given a score of 1 regardless of whether any Sargassum, lamentous algae or sessile invertebrates were present. If the substratum was generally at with few bumps its score was 3, while a very complex substratum prole scored 5. A score of 3 was assigned to substrata of intermediate complexity. Two refuge-type categories (holes and crevices) and three size ranges (\30 cm; 30 cm 1 m; [1 m) were used following the method proposed by Aburto-Opereza & Balart (2001). Holes in rocky shores were easily measured and counted, and crevices (gaps between structures that could provide a path for a sh to escape a predator) were estimated visually. Substratum complexity was also assessed by quantifying rock diversity in different size categories (\30 cm; 30 cm1 m; [1 m). Habitats that had rocks in a single size category scored 1 while substratum with all size categories present scored 5. The aim of this score was to assess the diversity of structural attributes across habitats, which have the potential to

provide settlement areas for benthic species and resources for different sh species. Habitat architecture height was a subjective visual estimate to distinguish between taller ([100 cm) and shorter (\50 cm) rocks in relation to the lowest point in the site. The height of the habitat architecture was used as a surrogate for the surface available for benthic colonisation. Hard substratum referred to the percentage of substratum that was not sand, rubble or shell sand patches. Physical structure was assessed using an index score adapted from Gratwicke & Speight (2005) (Table 1). A total score was calculated by adding the scores of each of the ve physical descriptors to give an estimate of the overall degree of complexity of the sites physical structure (Table 1). Sites that had a high physical structure had a total score equal or greater than 15, and sites with a low physical structure had a total score below 15. The majority of physical structure classications made a priori were in agreement with the physical structure index results. Sampling Three samples were undertaken at each site during each sampling occasion, between 2006 and 2008. Site sampling was accomplished on three sampling occasions in each season, totalling 90 samples in the dry/ winter season and 84 in the wet/summer season, yielding a total of 174 samples. High structured sites close (30 samples), near (30) and far (30) from the outfall yielded 90 samples, and low structured sites

Table 1 Physical structure score to discriminate between high and low structured sites Physical structure score 1 Rugosity (visual topographic estimate of the substratum in each site) Number of refuge size categories: refuge (2): holes and crevices; Size classes (3): \30 cm; 30 cm1 m; [1 m) Substratum complexity (number of categories): rock size classes (3): \30 cm; 30 cm1 m; [1 m. Height (visual estimate of height of physical structurecm) Hard substratum (%) Flat areasregardless of any branched algae or soft corals growing on it 01 3 Generally at with few bumps 24 5 Very complex substratum prole 46


050 020

51100 2040

[100 [40

High structured sites we those that have a total score C15, while low structured sites have a total score \15


Hydrobiologia (2012) 684:161175


close (36), near (18) and far (30) yielded 84 samples. Some replicates were not obtained as a result of inclement weather. The factors Season and Year were not included in the design, because according to our observations, rocky sh in this area did not change structure seasonally (Linear mixed models with randomised sites for the Number of individuals, F1,173 = 0.08, P = 0.77; Species richness, F1,173 = 0.008, P = 0.93) and there were not enough samples to perform comparisons by year. Underwater visual censuses were performed by SCUBA diving and snorkelling during a census parallel to the coastline along transects 30 m long and 3 m wide (90 m2). To obtain sh information, swimming along transects was performed twice. The rst time, the observer recorded the conspicuous species, and the second time the observer focused the search beneath rocks and in all crevices to observe the more cryptic species following the procedure of Aburto-Opereza & Balart (2001). The sampling unit, number of sh per 90 m2, was dened as the pooled number of conspicuous and cryptic species. Samples were performed in good weather and stable oceanographic conditions, between 9:00 and 14:00 h, during neap tide, near quarter moon. Additionally, we measured the percent of benthic cover from the 30 photographs taken for each site. Each photo was analysed for percentage cover using Count Point Coral with Excel ExtensionsCPCe 3.4 (Kohler & Gill, 2006), a software programme capable of random point analysis on digital photography. Twenty random points were generated for the photographic analysis. Benthic cover, similar to the physical structure analysis, was assessed based on photographs to quantify the percentage of each benthic class. Benthic cover descriptors comprised three algae divisions (Chlorophyta, Phaeophyta and Rhodophyta) and sessile invertebrates, mainly represented by the phyla Porifera, Mollusca (genus Petaloconchus vermitid), Echinodermata (Crinoid) and the class Anthozoa. The class Ascidiacea from the phylum Urochordata was also present. Benthic organisms, expressed as percentage of benthic cover, were grouped in the following classes (adapted from Gratwicke & Speight, 2005): (1) Growth forms applied very generally to living organisms such as coral, algae and other invertebrates that contribute to habitat complexity and support a variety of organism forms, such as branched, cylindrical, tube and pinnate among others.

With this class we assessed the diversity of structural attributes in habitats that could potentially provide resources for different sh species; (2) Encrusting cover referred to individuals that were encrusted in the substratum but did not form a complex habitat; (3) Sargassum spp., common and abundant pheaophyta in tropical rocky shores. (4) Palythoa caribaeorum encrusting coral; (5) Jania sp. rhodophyta algae; and (6) Petaloconchus sp. vermitid mollusc. During each sampling occasion, sub-surface (30 cm) and bottom (near to the bottom) water temperature were measured in triplicate. Data analyses Two-way analysis of variance was used to compare water temperature among the factor distances from the outfall and surface versus bottom, followed by a Tukey post hoc test. A principal component analysis (PCA) was performed on habitat structure descriptors (physical structure ? benthic cover) to detect site environmental patterns and to assess thermal inuence on benthic cover. Habitat structure descriptors included in the PCA were the scores obtained from physical structure index (rugosity, refuge size categories, substratum complexity, height and hard substratum) and benthic cover classes (growth forms, encrusting cover, Sargassum spp., Palythoa caribaerum, Jania sp. and Petaloconchus sp.). Benthic cover descriptors expressed as percentages were arcsinesquare-root transformed, and then a log10(x ? 1) transformation was applied to the whole dataset. Component loads greater than 0.5 were used to identify latent patterns. Two factors were tested using linear mixed models: Distance from the thermal discharge (close vs. near vs. far), 3 levels, and Physical structure (high vs. low), 2 levels. We compared sh species richness, number of individuals, Shannon diversity index (Magurran, 1988), densities of the 10 most abundant species and benthic cover classes among the distances from the outfall and physical structure (xed factors), with the sites included as a random factor. Linear mixed models take into account the nested structure of data, considering physical structure nested within the distance from the outfall, which are nested into sites. In addition, a Tukey post hoc test was performed when Ho was rejected. Prior to analysis, the data were transformed to stabilise variances and to minimise the



Hydrobiologia (2012) 684:161175

effect of abundant species (Underwood, 1997). Fish data were log10(x ? 1) transformed (Sokal & Rohlf, 1998). To explore the relationship between benthic cover (independent variables) and sh abundance (dependent variables), a stepwise multiple regression analysis was performed. Two-way analysis of similarity (ANOSIM) was used to test for signicant differences in sh community structure among distance from the discharges and physical structure. ANOSIM provides an R statistic that reects the amount of dissimilarity associated with each factor. An R value close to one indicates very different composition, whereas values near zero show little difference (Clarke & Warwick, 1994). To reveal the percentage contribution of each taxon to average similarity within each factor, we ran the similarity percentages (SIMPER). Analyses were based on the Bray-Curtis measure and performed with the computer programme PRIMER, version 5 (Clarke & Warwick, 1994).

Results Temperature decreased signicantly from the close to the far sites (F = 32.7, P \ 0.001), and surface temperature was signicantly different from bottom temperature (F = 24.3, P \ 0.001) at the close and near sites but not at the far sites (Fig. 2). Signicant distance versus water column interactions (F = 3.2,


Temperature ( oC)


bc b cd d


Close Near Far

Fig. 2 Box and whisker plot of variation in surface (white boxes) and bottom (black boxes) water temperatures among distance from the outfall categories. Letters indicate signicant difference levels from ANOVA at P \ 0.05

P = 0.04) were found although these were comparatively lower than the main effects of distance from outfall and depth (bottom versus surface). The highest temperature was 36.2C, recorded at the surface of the close sites, and the lowest temperature was 23.9C, recorded at the bottom of near and far sites. All benthic cover classes changed signicantly with distance to the thermal discharge. Branched calcareous algae from the Jania genus (F = 21.8, P \ 0.01) and Petaloconchus sp. vermitid (F = 175.4, P \ 0.01) were found almost exclusively at the sites close to the outfall. Jania species were most abundant at low structured sites (F = 20.6, P \ 0.01) and the Petaloconchus sp. vermitid was most abundant at high structured sites (F = 73.7, P \ 0.01). Growth forms were most abundant at sites near the outfall (F = 29.6, P \ 0.01), with decreasing abundance at further sites. Sargassum spp. were more abundant at the far sites compared to near and close sites (F = 81.5, P \ 0.01). Growth forms were signicantly more abundant at high structured sites (F = 6.1, P \ 0.05), while Sargassum spp. did not differ in abundance between high and low structured sites. Encrusting cover was widely distributed at all sites irrespective of distance from the outfall but was more abundant at low structured sites (F = 27.9, P \ 0.01) close and near to the outfall (F = 23.4, P \ 0.01), while Palythoa caribaeorum was found exclusively at the high structured sites near and far from the outfall (Fig. 3). The PCA showed that the high structured sites near and far from the outfall have more complex habitats compared to the low structured sites, and to a lesser extent, to the close high structured sites. The rst two PCA axes accounted for 43.8% (Axis 1, 26.9%; Axis 2, 16.8%) of the variance among sites. Axis 1 was positively correlated with rugosity and height and negatively correlated with encrusting cover and Jania sp. (Fig. 4). Axis 2 was positively correlated with Petalochonchus sp. and negatively correlated with growth forms and hard substratum. Sites highly structured close to the thermal outfall corresponded to high percentages of Petaloconchus sp., while low structured sites close to the thermal outfall corresponded to high percentage of Jania sp. (Fig. 4). On the other hand, high structured sites near and far from the thermal outfall corresponded to high rugosity, hard substrate and height. The low structured sites near and far from the thermal efuent had high percentages of encrusting cover and Sargassum spp.


Hydrobiologia (2012) 684:161175

30 20 10 0 60 40 20 0 60 40 20 0 60 40 20 0 60 40 20 0 40 20 0


Jania sp.

Petaloconchus sp.

Encrusting cover

Growth forms

Sargassum spp.

Palythoa caribaeorum




Fig. 3 Mean percent of benthic cover (SE) among the thermal gradient (close, near, far) and physical structure (white boxes low, black boxes high)

A total of 57 sh species comprising 32 families was observed. A member of the Haemulidae family (Haemulon steindachneri), a common group of coastal
Fig. 4 Distribution of sampling points on the rst two axes of the PCA based on habitat structure descriptors (physical ? benthic cover) according to distance from outfall (open circle close, open triangle near, open square far) and physical structure (white low, black high)
-2 3

sh, was the most abundant species ([9 individuals 9 90 m-2) and was widely distributed across all sites. The 20 most abundant species accounted for 97% of the total number of individuals and had an occurrence higher than 1% throughout all sites. The ve most abundant species at close and near sites accounted for 89% of the total number of individuals (each with a frequency of occurrence[4%) and were, in decreasing order: the gerreid Eucinostomus argenteus, the haemulid H. steindachneri, the pomacentrid A. saxatilis, the haemulid Haemulon aurolineatum, and the mugilid M. curema. Conversely, the ve most abundant species at sites far from the outfall, which contributed to 81% of the total number of sh (each with a frequency of occurrence [4%) were, in decreasing order: A. saxatilis, H. steindachneri, H. aurolineatum, the sparid Diplodus argenteus and the territorialist pomacentrid Stegastes fuscus. Species richness increased signicantly with distance from the outfall (F = 15.8, P \ 0.01) with higher values at near and far sites compared with the close sites. Moreover, the high structured sites had higher species richness (F = 9.8, P \ 0.01) compared with the low structured sites (Fig. 5a). The mean number of individuals (density) did not differ according to both the distance from the outfall and physical structure (Fig. 5b). Similarly to species richness, the Shannon diversity index differed among distance (F = 9.1, P \ 0.01) and physical structure (F = 5.1, P \ 0.01) (Fig. 5c).

Benthic cover (%)

Petaloconchus sp.


2,5 2 1,5 1
Jania sp. Substratum complexity

Refuge size categories Sargassum spp.

Height Rugosity

-1 -0,5
Encrusting cover

0 -1

Palythoa caribaeorum

Hard substratum

-1,5 -2

Growth forms



Hydrobiologia (2012) 684:161175

12 10


8 6 4 2 0

Individuals x 90m

150 125 100 75 50 25 0

Shannon (H')




Close Near Far

Fig. 5 Mean (SE) of sh species richness, number of individuals and Shannon diversity index among the thermal gradient (close, near, far) and physical structure (white low, black high)

Nine (E. argenteus, M. curema, S. greeleyi, A. saxatilis, D. argenteus, S. fuscus, H. aurolineatum, C. striatus and Malacoctenus delalandii) of the most abundant species differed signicantly among the distances from the outfall and six (M. curema, A. saxatilis, S. fuscus, H. aurolineatum, C. striatus and M. delalandii) differed between physical structure. E. argenteus (F = 5.1, P \ 0.01) and M. curema (F = 7.1; P \ 0.05) had higher abundances at the close sites compared with near and far sites. The puffersh Sphoeroides greeleyi had higher abundance at the sites close to and near the outfall (F = 17.7, P \ 0.01) compared with the far sites. In contrast, A. saxatilis (F = 33.3, P \ 0.01), D. argenteus, (F = 9.8, P \ 0.01), S. fuscus (F = 66.0, P \ 0.01), the butterysh Chaetodon striatus (F = 36.4, P \ 0.01) and the cryptic scaled blenny Malacoctenus delalandii (F = 29.1, P \ 0.01) were signicantly more abundant at sites far from the outfall. H. aurolineatum was the only species to have signicantly higher abundance at the near sites (F = 6.2, P = 0.01).

Abudefduf saxatilis (F = 10.3, P \ 0.05), S. fuscus (F = 82.1, P \ 0.01), C. striatus (F = 10.9, P \ 0.01) and M. delalandii (F = 23.2, P \ 0.01) were signicantly more abundant at the high structured sites, while H. aurolineatum (F = 13.6, P \ 0.01) was more abundant at the low structured sites (Fig. 6). When exploring the relationship between habitat structure (physical structure and benthic cover), species richness, number of individuals, Shannon index and abundance of the 10 most numerous species (Table 2), we found that with the exception of number of individuals and abundance of H. steindachneri, all sh variables showed signicant dependency on habitat structure, save for one species (H. aurolineatum) in which the proportion of variance explained was below 10%. The multiple regression analysis of species richness and diversity index, H0 , accounted for 26 and 19% of the variation of their estimated values, respectively. Twenty-six percent of total variation in species richness was explained by the variation in Petaloconchus sp., Jania sp. (negative relationship) and Palythoa caribaeorum and growth forms (positive relationship) (Table 2). Species abundance was generally higher in those sites with more P. caribaeorum and Sargassum spp. cover (e.g. A. saxatilis, C. striatus, M. delalandii and S. fuscus). In the case of E. argenteus and M. curema (positive relationship) and A. saxatilis (negative relationship), Petaloconchus sp. explained most of the variance. Signicant differences in the sh community among the distances from the thermal outfall were detected by ANOSIM (R global = 0.427; P \ 0.01), indicating shifts in the structure of the community. As expected, the largest difference was found between the close and far sites (R = 0.648), but signicant differences, although of smaller magnitude, were also found between the close and near (R = 0.366; P \ 0.01) and between the near and far (R = 0.209; P \ 0.01) sites. Furthermore, a low signicant difference (R global = 0.101; P \ 0.01) was also found between the high and low structured sites. An increased number of typical species was detected at the far sites from the thermal outfall (8 species), whereas the sites close to the outfall were characterised by four species only (Table 3). The low structured sites had a lower number of species that contributed to mean within-group similarity than the high structured sites.


Hydrobiologia (2012) 684:161175

Low High
100 10 1 0,1 100 10 1 0,1 100 10 1 0,1 100 10 1 0,1 100 10 1 0,1 100 10


E. argenteus

M. curema

S. greeleyi

H. steidacneri

H. aurolineatum

M. delalandii

0,1 100 10

C. striatus

0,1 100 10 1 0,1 100 10 1 0,1 100 10 1 0,1

S. fuscus

D. argenteus

A. saxatilis




Fig. 6 Mean abundance for the ten most numerous species (SE) along the thermal gradient (close, near, far) and physical structure (white low, black high)

Discussion Thermal discharges from the Brazilian cooling water nuclear power plants alter sh composition and habitat structure, and this factor was more inuential on sh assemblages than the physical structure of surrounding

habitats. Three species were directly associated with the most thermally impacted sites (close to and near the outfall) namely, E. argenteus, M. curema and S. greeleyi, but only the latter is a true rocky shore species. E. argenteus was the species that tolerates the most thermal stress, because it was highly abundant at sites that were close to the outfall. Our nding that species were less abundant at sites under thermal pollution coincides with the pattern reported by Rong-Quen et al. (2001), who conducted a 21-year study (19791999) on reef sh communities near a nuclear power station in southern Taiwan. Mora & na (2001) determined the critical thermal maxiOsp mum of reef sh from the tropical eastern Pacic and found that some species are able to colonise high temperature habitats, such as Haemulon steindachneri (36C), Mugil curema (40.8C) and Eucinostomus gracilis (38C). In this study, a similar tolerance pattern was found for Haemulon steindachneri, M. curema and a Eucinostomus species (not E. gracilis). These species had a high occurrence at impacted sites and were abundant mainly at those sites, suggesting a preference for areas with high temperatures. The abundant E. argenteus and M. curema have been reported mostly in sandy banks and muddy at areas, respectively (Benetti & Neto, 1991; Alvarez-Lajonchere et al., 1992; Chaves & Otto, 1999; Gaelzer & Zalmon, 2003). These species are probably already adapted to living in high temperature areas and as a consequence, they can use the resources available in these areas, whereas other species cannot subsist in high temperatures at all. As a result of living in areas impacted by thermal inuence, sh can alter their behaviour. For instance, some studies on thermal inuence have focused on aspects of reproductive biology. Increased temperature can have positive or negative effects on reproductive capacity depending on whether the species is near its optimal thermal point (Munday et al., 2008). Ruttenberg et al. (2005) found that egg production in Stegastes beebei is optimal at 25C and decreases to a very low production rate at either 20 or 27C. This suggests that with only a few degrees of temperature difference the reproductive success of a given species can signicantly decrease. One possible strategy for circumventing this problem is to change the timing of reproduction. Some information on the inuence of thermal discharge on freshwater sh from temperate areas is

Individuals x 90m



Hydrobiologia (2012) 684:161175

Table 2 Summary of the results of multiple linear regression analyses for sh assemblage structure and abundance of the 10 most numerous species, indicating the benthic cover independent variables included in each model and the sign of their relationship Variable Species richness Adj. R2 0.26 P *** Variables included in the model - Petaloconchus sp.** - Jania sp.* ? Palythoa caribaeorum** ? Growth forms** Number of individuals Shannon diversity index Eucinostomus argenteus Mugil curema Sphoeroides greeleyi Abudefduf saxatilis 0.03 0.19 0.37 0.18 0.19 0.30 * *** *** ** ** *** - Jania sp.* ? Palythoa caribaeorum** ? Growth forms* ? Petaloconchus sp.** ? Encrusting cover* ? Petaloconchus sp.** - Sargassum spp.** - Palythoa caribaeorum** - Petaloconchus sp.* ? Palythoa caribaeorum** ? Sargassum spp.** ? Growth forms** Haemulon steindachneri Diplodus argenteus Stegastes fuscus 0.04 0.20 0.38 * ** ** ? Palythoa caribaeorum** ? Growth forms** ? Sargassum spp.** - Petaloconchus sp.** ? Palythoa caribaeorum** ? Sargassum spp.** ? Growth forms** ? Encrusting cover** Haemulon aurolineatum Chaetodon striatus 0.07 0.17 ** *** - Petaloconchus sp.* ? Sargassum spp.* ? Sargassum spp.** ? Palythoa caribaeorum** ? Growth forms* Malacoctenus delalandii 0.18 ** ? Encrusting cover* ? Palythoa caribaeorum** ? Sargassum spp.** ns not signicant; * P \ 0.05, ** P \ 0.01, *** P \ 0.001

m (1994) found that available. Luksiene & Sandstro Rutilus rutilus fail to recruit in areas that are exposed to thermal discharge from a nuclear power station in Sweden. Madenjian et al. (1986) found a decreased abundance of Alosa pseudoharengus and Perca avescens at a nuclear power plant on Lake Michigan when compared to a reference condition. Negative effects of thermal inuence on the growth and the spawning period of Micropterus dolomieu were found Bay in Lake Huron (Shuter et al. 1985). for the Dore

We also found that physical structure plays an important role in sh species composition and distribution along the thermal gradient that seems to have more relevance at sites far from the thermal outfall. For example, higher occurrences and abundances of S. fuscus, A. saxatilis, C. striatus and M. delalandii were found only at high structured sites far from the outfall. This can be attributed to the high physical structure and benthic cover found in these areas. These species are cryptic or territorial and habitat complexity is an


Hydrobiologia (2012) 684:161175


Table 3 Contribution made by each species to the overall measure of similarity within the factor (distance from the outfall and physical structure) Distance from the outfall Similarity average (%) Close 37.49 56.94 17.82 10.15 6.85 25.53 5.50 4.74 2.83 7.76 4.17 3.56 2.11 7.84 40.98 8.92 6.20 18.41 9.64 Near 40.10 11.64 9.25 32.96 18.03 Far 34.31 Physical structure Low 27.39 29.45 10.14 22.65 High 30.93 12.49 8.31 24.02 3.27 31.35 3.00

Eucinostomus argenteus Sphoeroides greeleyi Haemulon steindachneri Mugil curema Abudefduf saxatilis Diplodus argenteus Haemulon aurolineatum Serranus aviventris Stegastes fuscus Chaetodon striatus Malacoctenus delalandii Halichoeres poeyi

Only species contributing to more than 2% similarity according to SIMPER are shown

important factor providing shelters, feeding and reproductive grounds (Ferreira et al., 2001). In contrast, warm-water species such as E. argenteus, M. curema and S. greeleyi, and species that occurred mainly at far sites, such as D. argenteus, seemed to not be inuenced by physical structure. Physical structure is closely related to sh species richness (Letourneur, 1996; Friedlander & Parrish, 1998; Grober-Dunsmore et al., 2008) because it provides shelter opportunities and places for sh to hman & Rajasuriya, reproduce. Several studies (O a1998; Aburto-Opereza & Balart, 2001; Garc Charton et al., 2004; Tittensor et al., 2007) have reported the effects of habitat structure on sh communities, but most of them fail to discriminate the particular inuence of either physical structure or benthic cover. In this study, sites classied as highly structured had more structural complexity and showed greater rugosity and height than the other sites. They were comprised of more physical structures, such as rocks of different sizes, holes and crevices. These features were closely related to higher sh richness and abundance as compared to the low structured sites. Structural complexity per se, aside from the benets that structure can provide such as shelter or increased surface area for accumulation of food, may not be greatly attractive to juvenile sh (Laegdsgaard & Johnson, 2001). However, these benets can be

important for habitat selection for certain species. Cabaitan et al. (2008) found that adding live corals to consolidated dead corals increased juvenile sh colonisation, with these patches acting as true articial reefs to concentrate organisms. Friedlander & Parrish (1998) reported that physical structure and species abundance are weakly correlated, while Gladfelter et al. (1980) suggested that the abundance of coral sh increases with structural complexity. Therefore, physical structure, when coupled with diverse benthic cover, provided the best conguration for supporting a rich and diverse sh community, as it was observed at the high structured sites far from the outfall. High sh richness was found at high structured sites near to the outfall. This suggests that the thermal plume was restricted to the more supercial layers. A small difference (about 2C) between surface and bottom water temperatures seemed not to affect benthic cover, and consequently, sh species richness, which was favoured by high habitat complexity. Thermal discharge affected habitat structure more intensely at the sites close to the outfall, which had low benthic cover. In these sites, a few organisms were found, such as vermetid molluscs of the Petaloconchus genus and calcareous algae of the Jania genus. Therefore, low richness at the most impacted sites was associated with thermal inuence, and consequently, low benthic cover, which limited feeding resource



Hydrobiologia (2012) 684:161175

availability. Furthermore, sh abundance was high at the sites close to the outfall, indicating that tolerant species can take advantage of such unoccupied niches and increase in number. Overall, to support high sh richness and abundance, high benthic cover and large rocky areas are required; a notion corroborated at the high structured sites near and far from the outfall. Lardicci et al. (1999), studying thermal discharges on the spatial distribution of meiobenthic and macrobenthic communities in Follonica Gulf (Western Mediterranean), found that thermal efuent did not inuence benthic community distribution. On the other hand, Vilanova et al. (2004), working in the same area as this study, found differences in richness and abundance of sponge communities between an area under thermal inuence from the BNPP and a control area. Overall, sessile benthic organisms have been recorded as being susceptible to thermal efuent, and an increase by a few degrees can jeopardise their survival (Laws, 1993; Logue et al., 1995), as was observed at the sites close to the outfall. This suggests that thermal pollution is the main limiting factor for the composition and abundance of benthic communities. In this study, most of the algae species were incrustant and lamentous with limited capacity to form banks and complex patches. Even the growth forms species had a small size (height average = 15 cm, T. P. Teixeira personal observation) with low contribution to structural complexity. Vegetated areas are attractive for sh because they are used as feeding resources and refuge against predators. Conversely, unvegetated areas usually have low sh abundance and diversity (Robertson, 1984; Bell et al., 1987; Ornellas & Coutinho, 1998). Several plants are associated with high diversity of invertebrates, and they are an important part of the diet of juvenile sh (Lubbers et al., 1990; Schneider & Mann, 1991). Zalmon et al. (2002), working with articial reefs, detected that increased species richness of the reef sh community may be related to the gradual development of a fouling community on the rocky structure. In this study, we conrmed the inuence of thermal pollution on the biota. However, little information is available on the effects of the damage caused by high temperatures on the distribution of sh in tropical areas, and the present ndings can provide relevant information for mitigating the potential impacts caused by this kind of anthropogenic inuence that is becoming very common in developing countries.

Other natural environmental stressors not accounted for in this study (e.g. predation, limitation in food availability) and human disturbance (e.g. sheries, pollution, increased ow near to the outfall, water chlorine discharges), may also play an important role in structuring sh communities. For instance, chlorine is added to the power station cooling water for antibiotic purposes (Jenner et al., 1997), and chlorination by-products (CBPs) may act as biocides once the water is introduced into the ocean. Overall, the sites close to the thermal discharge were more heavily impacted by thermal pollution, reected by low benthic cover, and consequently, low sh richness, where opportunist and tolerant species dominated. On the other hand, areas that were less impacted were colonised by typical rocky shore species. The more structured sites minimised the thermal inuence, as was observed at some sites near the thermal discharge that had benthic cover and a sh community that was relatively unaffected. We have provided basic information for environmental managers to minimise the inuence of thermal pollution and to contribute to the knowledge of changes in habitat structure and sh communities in thermally impacted coastal tropical areas. Environmental policies should consider that the effects of cooling water from nuclear power plants can be minimised by increasing the complexity of physical structures at some hundreds of metres away from the discharges, where, as we report in the study, benthic cover can develop. In such cases, the thermal plume may be restricted to the surface layers making the thermal effects less harmful for sh communities.
Acknowledgments We thank Hamilton Hissa Pereira and Rafael Jardim Albieri for their help in the eld work, and to technical staff of the Laboratory of Fish Ecology, University Federal Rural of Rio de Janeiro for useful help in the laboratory. This study was partially nanced by CNPqBrazilian National Council for Research Development (Proc. 302555/2008-0).

Aburto-Opereza, O. & E. F. Balart, 2001. Community structure of reef sh in several habitats of a rocky reef in the Gulf of California. Marine Ecology 22: 283305. Alvarez-Lajonchere, L., E. Trewavas & G. J. Howes, 1992. Mugil curema and Mugil liza Valenciennes in Cuvier and Valenciennes, 1836 (Osteichthyes, Perciformes): proposed conservation of the specic names. Bulletin of Zoological Nomenclature 49: 271275.


Hydrobiologia (2012) 684:161175 Bamber, R. N. & J. F. Spencer, 1984. The benthos of a coastal power station thermal discharge canal. Journal of the Marine Biological Association of the United Kingdom 64: 603623. Bandeira, J. V., A. A. Barreto, V. L. Bomtempo, R. M. Moreira & L. H. Salim, 2003. The performance of a surface jet on a closed bay: the reality compared with the theory. Proceedings of 6th international conference on coastal and port engineering in developing countries, Colombo, 15 September 2003, CD-ROM, COPEDEC VI, Colombo, Sri Lanka Beitinger, T. L., W. A. Bennett & R. W. McCauley, 2000. Temperature tolerances of North American freshwater shes exposed to dynamic changes in temperature. Environmental Biology of Fishes 58: 237275. Bell, J. D., M. Westoby & A. S. Steffe, 1987. Fish larvae settling in seagrass. Do they discriminate between bed of different leaf density? Journal of Experimental Marine Biology and Ecology 111: 133134. Benetti, D. D. & E. B. F. Neto, 1991. Preliminary results on growth of mullets (Mugil liza and Mugil curema) fed articial diets. World Aquaculture 22: 5557. Bennett, W. A. & F. W. Judd, 1992. Comparison of methods for determining low temperature tolerance: experiments with pinsh, Lagodon rhomboids. Copeia 1992: 10591065. Bruno, J. F., E. R. Selig, K. S. Casey, C. A. Page, B. L. Willis, C. D. Harvell, H. Sweatman & A. M. Melendy, 2007. Thermal stress and coral cover as drivers of coral disease outbreaks. Public Library of Science (PLoS) Biology 5(6): e124. o, 2008. Effects of Cabaitan, P. C., E. D. Gomez & P. M. Alin coral transplantation and giant clam restocking on the structure of sh communities on degraded patch reefs. Journal of Experimental Marine Biology and Ecology 357: 8598. Chabanet, P., H. Ralambondrainy, M. Amanieu, G. Faure & R. Galzin, 1997. Relationships between coral reef substrata and sh. Coral Reefs 16: 93102. Chapman, M. G., A. J. Underwood & G. A. Skilleter, 1995. Variability at different spatial scales between a subtidal assemblage exposed to the discharge of sewage and two control assemblages. Journal of Experimental Marine Biology and Ecology 189: 103122. Chaves, P. T. C. & G. Otto, 1999. The mangrove as a temporary habitat for sh: the Eucinostomus species at Guaratuba Bay, Brazil (25520 S; 48390 W). Brazilian Archives of Biology and Technology 42: 6168. Chou, Y., T. Y. Lin, C. T. A. Chen & L. L. Liu, 2004. Effect of nuclear power plant thermal efuent on marine sessile invertebrate communities in Southern Taiwan. Journal of Marine Science and Technology 12: 448452. Clarke, K. R. & R. M. Warwick, 1994. Change in Marine Communities: An Approach to Statistical Analysis and Interpretation. Natural Environment Research Council, Plymouth. Contador, J. F. L., 2005. Adaptive management, monitoring, and the ecological sustainability of a thermal-polluted water ecosystem: a case in SW Spain. Environmental Monitoring and Assessment 104: 1935. De Vries, P., J. E. Tamis, A. J. Murk & M. G. D. Smit, 2008. Development and application of a species sensitivity distribution for temperature-induced mortality in the aquatic

173 environment. Environmental Toxicology and Chemistry 27: 25912598. Dembski, S., G. Masson, D. Monnier, P. Wagner & J. C. Pihan, 2006. Consequences of elevated temperatures of life-history traits of an introduced sh, pumpkinseed Lepomis gibbosus. Journal of Fish Biology 69: 331346. Devinny, J. S., 1980. Effects of thermal efuents on communities of benthic marine macro-algae. Journal of Environmental Management 11: 225242. guez-Ruiz & C. Granado-Lorencio, 2008. Encina, L., A. Rodr Distribution of common carpina Spanish reservoir in relation to thermal loading from a nuclear power plant. Journal of Thermal Biology 33: 444450. Ferreira, C. E. L., J. E. A. Gonc alves & R. Coutinho, 2001. Community structure of shes and habitat complexity on a tropical rocky shore. Environmental Biology of Fishes 61: 353369. Friedlander, A. M. & J. D. Parrish, 1998. Habitat characteristics affecting sh assemblages on a Hawaiian coral reef. Journal of Experimental Marine Biology and Ecology 26: 6584. Gaelzer, L. R. & I. R. Zalmon, 2003. The inuence of wave gradient on the ichthyofauna of Southeastern Brazil: focusing the community structure in surf-zone. Journal of Coastal Research 35: 456462. a-Charton, J. A., A. Pe rez-Ruzafa, P. Sa nchez-Jerez, J. Garc ones & D. Moreno, 2004. MultiT. Bayle-Sempere, O. Ren scale spatial heterogeneity, habitat structure, and the effect of marine reserves on Western Mediterranean rocky reef sh assemblages. Marine Biology 144: 161182. Gladfelter, W. B. & E. H. Gladfelter, 1978. Fish community structure as a function of habitat structure on West Indian patch reefs. Revista de Biologia Tropical 26: 6584. Gladfelter, W. B., J. C. Ogden & E. H. Gladfelter, 1980. Similarity and diversity among coral reef communities: a comparison between tropical western Atlantic (Virgin Islands) and tropical central Pacic (Marshall Islands) patch reefs. Ecology 61: 11561168. Gratwicke, B. & M. R. Speight, 2005. The relationship between sh species richness, abundance and habitat complexity in a range of shallow tropical marine habitats. Journal of Fish Biology 66: 650667. Grober-Dunsmore, R., T. K. Frazer, J. P. Beets, W. J. Lindberg, P. Zwick & N. A. Funicelli, 2008. Inuence of landscape structure on reef sh assemblages. Landscape Ecology 23: 3753. Jenner, H. A., C. J. L. Taylor, M. Van Donk & M. Khalanski, 1997. Chlorination by products in chlorinated cooling water of some European costal power station. Marine Environmental Research 43: 279293. Jokiel, P. L. & S. L. Coles, 1974. Effects of heated efuent on hermatypic corals at Kahe Point, Oahu. Pacic Science 28: 118. Kohler, K. E. & S. M. Gill, 2006. Coral Point Count with Excel extensions (CPCe): A Visual Basic program for the determination of coral and substrate coverage using random point count methodology. Computers and Geosciences 32: 12591269. Krishnakumar, V., J. S. Sastry & G. N. Swamy, 1991. Implication of thermal discharges into the seaa review. Indian Journal of Environmental Protection 11: 525527.


174 Laegdsgaard, P. & C. Johnson, 2001. Why do juvenile sh utilize mangrove habitats? Journal of Experimental Marine Biology and Ecology 257: 229253. Langford, T. E. L., 1990. Ecological Effects of Thermal Discharges. Elsevier, London. Lardicci, C., F. Rossi & F. Maltagliati, 1999. Detection of thermal pollution: variability of benthic communities at two different spatial scales in an area inuenced by a coastal power station. Marine Pollution Bulletin 38: 296303. Laws, E. A., 1993. Aquatic PollutionAn Introductory Text. Wiley, New York. Letourneur, Y., 1996. Dynamics of sh communities on Reunion fringing reefs, Indian Ocean: 1. Patterns of spatial distribution. Journal of Experimental Marine Biology and Ecology 195: 130. Logue, J., P. Tiku & A. R. Cossins, 1995. Heat injury and resistance adaptation in sh. Journal of Thermal Biology 20: 191197. Lubbers, L., W. R. Boynton & W. M. Kemp, 1990. Variations in structure of estuarine sh communities in relation to abundance of submersed vascular plants. Marine Ecology Progressive Series 65: 114. Lucca, E. V. D., J. V. Bandeira, J. A. Lorenzzetti, R. C. Moreira, nior & E. S. C. R. M. Castro, L. H. Salim, O. D. Z. Ju Esposito, 2005. Uso de sensor hiperespectral aerotrans nica portado no monitoramento da pluma termal ocea o da central nuclear decorrente da descarga de refrigerac a de Angra dos Reis. Revista Brasileira de Cartograa 57: 4855. m, 1994. Reproductive disturbance Luksiene, D. & O. Sandstro in a roach (Rutilus rutilus) population affected by cooling water discharge. Journal of Fish Biology 45: 13625. Madenjian, C. P., D. J. Jude & F. J. Tesar, 1986. Intervention analysis of power-plant impact on sh populations. Canadian Journal of Fisheries and Aquatic Sciences 43: 819829. Magurran, A. E., 1988. Ecological Diversity and Its Measurement. Croom Helm, London. Mahadevan, S., 1980. A study on the effects of power plant thermal discharges on benthic infaunal communities at Big Bend, Tampa Bay (Florida). Florida Science 43: 78. Mariazzi, A. A., J. L. Donadelli, P. Arenas, M. A. Di Siervi & C. Bonetto, 1992. Impact of a nuclear power plant on water quality of Embalse del Rio Tercero Reservoir, (Cordoba, Argentina). Hydrobiologia 246: 129140. McCormick, M. I., 1994. Comparison of eld methods for measuring surface topography and their associations with a tropical reef sh assemblage. Marine Ecology Progress Series 112: 8796. na, 2001. Tolerance to high temperatures Mora, C. & A. F. Osp and potential impact of sea warming on reef shes of Gorgona Island (tropical eastern Pacic). Marine Biology 139: 765769. Munday, P. L., G. P. Jones, M. S. Pratchett & A. J. Williams, 2008. Climate change and the future for coral reef shes. Fish and Fisheries 9: 261285. hman, M. C. & A. Rajasuriya, 1998. Relationships between O habitat structure and sh communities on coral and sandstone reefs. Environmental Biology of Fishes 49: 4561.

Hydrobiologia (2012) 684:161175 Ornellas, A. B. & R. Coutinho, 1998. Spatial and temporal patterns of distribution and abundance of a tropical sh assemblage in a seasonal Sargassum bed, Cabo Frio Island, Brazil. Journal of Fish Biology 53(Suppl A): 198208. Ospina, A. F. & C. Mora, 2004. Effect of body size on reef sh tolerance to extreme low and high temperatures. Environmental Biology of Fish 70: 339343. Penaz, M., V. Barus & M. Prokes, 1999. Changes in the structure of sh assemblages in a river used for energy production. Regulated Rivers-Research & Management 15: 169180. Qian, S., H. Chen, X. Zhao & Q. Zhang, 1993. A study of the effect of thermal pollution on seaweeds. Journal Ocean University of Qingdao 23: 2234. Roberts, C. M. & R. F. G. Ormond, 1987. Habitat complexity and coral reef sh diversity and abundance on Red Sea fringing reefs. Marine Ecology Progress Series 41: 18. Robertson, A. I., 1984. Trophic interactions between the sh fauna and macrobenthos of an eelgrass community in Western Port, Victoria. Aquatic Botany 18: 135153. Rong-Quen, J., C. Jeng-Ping, L. Chun-Yu & S. Kwang-Tsao, 2001. Long-term monitoring of the coral reef sh communities around a nuclear power plant. Aquatic Ecology 35: 233243. Ruttenberg, B. I., A. J. Haupt, A. I. Chiriboga & R. R. Warner, 2005. Patterns, causes and consequences of regional variation in the ecology and life history of a reef sh. Oecologia 145: 394403. Sandstrom, O., I. Abrahamsson, J. Andersson & M. Vetemaa, 1997. Temperature effects on spawning and egg development in Eurasian perch. Journal Fish Biology 51: 10151024. Schneider, F. I. & K. H. Mann, 1991. Species specic relationships of invertebrates to vegetation in a seagrass bed. I, Correlational studies. Journal of Experimental Marine Biology and Ecology 145: 101117. Schubel, J. R., C. C. Coutant & P. M. J. Woodhead, 1978. Thermal effects of entrainment. In Schubel, J. R. & B. C. Marcy (eds), Power Plant EntrainmentA Biological Assessment. Academic Press Inc., New York: 1993. Shuter, B. J., D. A. Wismer, H. A. Regier & J. E. Matuszek, 1985. An application of ecological modeling: impact of thermal efuent on a smallmouth bass population. Transactions of the American Fisheries Society 114: 63651. Sokal, R. R. & J. Rohlf, 1998. Biometry: The Principles and Practice of Statistics in Biological Research. W.H. Freeman, New York. Suresh, K., M. S. Ahamed, G. Durairaj & K. V. K. Nair, 1993. Impact of power plant heated efuent on the abundance of sedentary organism, off Kalpakkam, east coast of India. Hydrobiologia 268: 109114. jo, 2009. Effects of a Teixeira, T. P., L. M. Neves & F. G. Arau nuclear power plant thermal discharge on habitat complexity and sh community structure in Ilha Grande Bay, Brazil. Marine Environmental Research 68: 188195. m & B. Worm, 2007. Tittensor, D. P., F. Micheli, M. Nystro Human impacts on the speciesarea relationship in reef sh assemblages. Ecological Letters 10: 760772. Underwood, A. J., 1997. Experiments in Ecology: Their Logical Design and Interpretation Using Analysis of Variance. Cambridge University Press, Cambridge.


Hydrobiologia (2012) 684:161175 Verlaque, M., G. Giraud & C. F. Boudouresque, 1981. Effects of a thermal power plant on the Mediterranean marine phythobenthos: the area of high frequency temperature changes. Botanica Marina 24: 6987. Verones, F., M. M. Hanaah, S. Pster, M. A. J. Huijbregts, G. J. Pelletier & A. Koehler, 2010. Characterization factors for thermal pollution in freshwater aquatic environments. Environmental Science & Technology 44: 93649369. Vilanova, E., M. M. Pinto, M. P. Curbelo-Fernandez & S. H. G. Silva, 2004. The impact of a nuclear power plant discharge on the sponge community of a tropical Bay (SE Brazil).

175 Bollettino dei Musei e degli Istituti Biologici dello di Genova 68: 647654. Universita Wright, J. F., J. M. R. Winder, J. M. Gunn, J. H. Blackburn, K. L. Symes & R. T. Clarke, 2000. Minor local effects of a river Thames power station on the macroinvertebrate fauna. Regulated Rivers: Research & Management 16: 159174. Zalmon, I. R., R. Novelli, M. P. Gomes & V. V. Faria, 2002. Experimental results of an articial reef programme on the Brazilian coast north of Rio de Janeiro. ICES Journal of Marine Science 59: 8387.